Diagnostic accuracy and cost-effectiveness of the CAR-OLT score in predicting cardiac risk for liver transplantation

Abstract

BACKGROUND

The CAR-OLT score predicts major adverse cardiovascular events 1 year after liver transplant (LT).

AIM

To test the hypothesis that the CAR-OLT score may help avoid cardiac stress tests in LT candidates.

METHODS

This retrospective single-center cohort study included all adult patients undergoing elective evaluation for first cadaveric donor orthotopic LT for liver cirrhosis with or without hepatocellular carcinoma at Fondazione Policlinico Universitario Agostino Gemelli Istituto di Ricerca e Cura a Carattere Scientifico in Rome, Italy. Cardiac contraindications for LT listing were defined after a center-specific cardiac workup, which included cardiac stress tests for most patients. The diagnostic accuracy of the CAR-OLT score was evaluated using the area under the receiver operating characteristic (AUROC) method.

RESULTS

A total of 342 LT candidates were evaluated between 2015 and 2019, with a moderate cardiovascular risk profile (37% diabetes, 34% hypertension, 22% obesity). Of these, 80 (23%) candidates underwent coronary angiography. Twenty-one (6%) candidates were given cardiac contraindications to LT listing, 48% of which were due to coronary artery disease. The CAR-OLT score predicted cardiac contraindications to LT listing with an AUROC of 0.81. The optimal cut-off for sensitivity was a CAR-OLT score ≤ 23, which showed a 99% negative predictive value for cardiac contraindications to LT listing. A total of 84 (25%) LT candidates with a CAR-OLT score ≤ 23 underwent 87 non-invasive cardiac tests and 13 coronary angiographies pre-listing, with estimated costs of approximately 48000€. The estimated savings per patient was €574.70 for the Italian National Health System.

CONCLUSION

A CAR-OLT score ≤ 23 can identify LT candidates who can be safely listed without the need for cardiac stress tests, providing time and cost savings. These findings require external validation.

Key Words: Transplantation; Coronary artery disease; Pharmacoeconomic; Listing; Major adverse cardiovascular events

Core Tip: The burden of cardiovascular risk factors and coronary artery disease among liver transplant (LT) candidates is increasing, but the optimal diagnostic cardiac work-up remains debated. The CAR-OLT score (available at www.carolt.us), based on twelve pre-transplant characteristics, predicts major adverse cardiovascular events one year after LT. In our population, a CAR-OLT score cut-off of ≤ 23 identifies low-risk patients who could directly proceed to LT listing without additional diagnostic tests. This approach could have avoided 87 stress tests and 13 coronary angiographies, saving an estimated €574.70 per patient for the Italian National Health System.

INTRODUCTION

Cardiovascular events are the leading cause of both early morbidity[1] and early mortality[2] after liver transplantation (LT), the third leading cause of late mortality, and are showing an upward trend[3,4]. The burden of cardiovascular risk factors and coronary artery disease (CAD) among LT candidates is rising for several reasons, including the increase in waiting list registrations of older candidates[5] and patients with metabolic dysfunction-associated steatohepatitis (MASH)[6-8]. Additionally, cirrhotic cardiomyopathy, a condition characterized by blunted inotropic and chronotropic responsiveness to stress, also plays a significant role in determining cardiovascular events after LT[9,10]. International guidelines recommend either non-invasive stress tests or anatomic CAD tests for LT candidates with multiple cardiovascular risk factors[11-14]. However, the proposed algorithms have not been validated, and the optimal cardiovascular assessment for LT candidates remains debated.

Three LT-specific cardiovascular risk scores have been published. In 2021, Rachwan et al[15] proposed the CAD-LT score, based on seven pre-LT characteristics, which effectively stratifies pre-LT risk for significant CAD and may help avoid unnecessary non-invasive or invasive diagnostic procedures[15]. Martinez-Perez et al[16] proposed stratifying cardiovascular risk among LT candidates based on the CAD-LT score, reserving coronary computed tomography (CT) angiography for candidates with a CAD-LT score ≥ 7. However, Pagano et al[17] observed that the modified CAD-LT score (excluding family history of CAD) tends to overestimate the presence of CAD in their cohort[17]. Furthermore, the CAD-LT score focuses solely on CAD, which accounts for less than 10% of major adverse cardiovascular events (MACE) after LT[18]. The Cardiac Arrest Risk Index score, which includes age, sex, model for end-stage liver disease (MELD) score, and QTc interval prolongation, may identify patients at higher risk of cardiac arrest or ventricular arrhythmia within 30 days post-LT[19].

In 2017, VanWagner et al[20] developed the CAR-OLT score to predict the 1-year risk of death or hospitalization due to a MACE after LT. The CAR-OLT score (available at www.carolt.us) is calculated based on twelve pre-LT characteristics collected from clinical history and does not include non-invasive stress test results. Recently, Tang et al[21] externally validated the CAR-OLT score in a cohort of 467 LT patients from Pennsylvania, showing suboptimal discrimination for 1-year MACE[21]. We tested the hypothesis that the CAR-OLT score can predict cardiac contraindications for LT listing and identify a group of low-risk patients who could directly proceed to LT listing, potentially avoiding unnecessary cardiac stress tests. The aim of this study is to evaluate the diagnostic accuracy of the CAR-OLT score in predicting the presence of cardiac contraindications in LT candidates.

MATERIALS AND METHODS

Inclusion/exclusion criteria and definitions

This retrospective cohort study included all adult patients (age ≥ 18 years) undergoing elective evaluation for first cadaveric donor orthotopic LT for liver cirrhosis, with or without hepatocellular carcinoma, at the LT center of the Fondazione Policlinico Universitario Agostino Gemelli Istituto di Ricerca e Cura a Carattere Scientifico in Rome, Italy, who had completed the preoperative cardiac evaluation between 2015 and 2019. Candidates for re-transplantation, double liver-kidney transplantation, acute liver failure, or those with an indication for LT other than liver cirrhosis (e.g., polycystic liver disease, biliary tract atresia, Caroli disease, Rendu-Osler disease, Budd-Chiari syndrome) were excluded. Clinical data were retrospectively collected from paper and electronic medical records within the hospital’s computer systems. Extracted data included clinical demographics, etiology of cirrhosis, cardiac risk factors, MELD score[22] at the time of listing and LT, non-invasive stress test results, and coronary angiography findings. Obesity was defined as a body mass index (BMI) ≥ 30[23]. Portal vein thrombosis was identified from pre-transplant imaging. A personal history of CAD was defined as a history of previous percutaneous coronary intervention, coronary artery bypass grafting, or myocardial infarction. Similarly, a family history of CAD was defined as the occurrence of any of the aforementioned conditions in a first-degree relative. A pulmonary arterial systolic pressure (PASP) value ≥ 40 mmHg, measured on a color-Doppler echocardiogram, was the threshold for referring patients to right heart catheterization[24]. An exercise electrocardiogram was considered valid if it reached at least 85% of the age-predicted maximum exercise heart rate[25]. Coronary stenosis was considered significant if the stenosis diameter was ≥ 70% for non-left main disease or ≥ 50% for left main disease[26].

Pre-transplant cardiac workup

Electrocardiograms and color doppler echocardiograms were performed on all patients. In the presence of significant findings on these tests (e.g., valvular diseases, elevated PASP), patients underwent cardiological evaluation and were managed accordingly. Non-invasive stress testing was required for all males older than 40 years, females older than 50 years, or individuals with one or more cardiovascular risk factors (arterial hypertension, diabetes mellitus, cigarette smoking, hypercholesterolemia, obesity, or a family history of CAD). Due to local availability, the first-choice test at our institution was stress myocardial perfusion imaging (preferably with physical exercise on a cycle ergometer, otherwise with pharmacological stress using dipyridamole or adenosine). Other non-invasive stress tests included exercise electrocardiograms, dobutamine stress echocardiography, and coronary CT angiography, depending on the recommendation of a dedicated cardiology consultant, who reviewed all candidates before LT listing. In case of a positive or inconclusive result after non-invasive stress tests, coronary angiography was indicated. If non-significant CAD was found, the patient was listed after optimization of medical therapy. If significant CAD was detected, the patient was evaluated by the local multidisciplinary heart team for a final decision on listing and myocardial revascularization. Coronary angiography was performed directly in patients with multiple cardiovascular risk factors for whom non-invasive stress tests could not be safely performed. Patients with PASP ≥ 40 mmHg on color doppler echocardiogram underwent right heart catheterization. Porto-pulmonary hypertension was defined as a mean pulmonary artery pressure > 25 mmHg, pulmonary artery wedge pressure < 15 mmHg, and pulmonary vascular resistance > 3 wood units. Our study protocol includes pulmonary function tests for all patients, except for those who are the most critically ill and unable to undergo the test.

Cardiac contraindications to liver transplant

At our institution, we consider the following conditions as prohibitive for accepting a patient as a LT candidate: (1) Non-revascularized severe or significant CAD; (2) Decreased left ventricular systolic function (ejection fraction < 45%) and/or moderate to severe right heart failure, not amenable to medical or percutaneous treatments; (3) Uncontrolled pulmonary hypertension, defined as mean pulmonary arterial pressure ≥ 45 mmHg despite maximal medical management; (4) Significant uncorrectable structural valvular abnormalities (e.g., aortic stenosis, mitral stenosis, aortic regurgitation, mitral regurgitation); and (5) Recurrent unstable arrhythmias.

Study outcomes

The primary outcome of the study was to evaluate the diagnostic accuracy of the CAR-OLT score in predicting the presence of cardiac contraindications to LT listing, as determined after a center-specific diagnostic cardiac workup, as previously reported.

Pharmacoeconomic analysis

The pharmacoeconomic analysis of the study was conducted by considering all invasive and non-invasive tests used to stratify the cardiovascular risk of LT candidates. The costs of these tests were calculated based on the current healthcare system's reimbursement rates. A scenario was then modeled in which the CAR-OLT test was applied to all candidates a priori, assuming that no further tests would be performed on patients with a CAR-OLT value below the cut-off identified in the study. Finally, potential savings were estimated by calculating the number of invasive and non-invasive tests that could have been avoided if a strategy based on the a priori application of the CAR-OLT score had been implemented.

The reimbursement rates for diagnostic tests were taken from the Lazio Regional Health System Outpatient Specialist Services Tariff Schedule 2022[27]. For coronary angiography, the rate corresponding to DRG 125 for a two-night ordinary hospitalization with procedure code 88.56 was used[28]. The cost estimates for each diagnostic test are as follows: Exercise electrocardiogram, 55.80€; myocardial perfusion imaging, 308€; stress echocardiogram, 70€; coronary CT angiography, 158€; coronary angiography, 2142€.

Statistical analysis

Descriptive statistics were calculated using mean, median, standard deviation, interquartile range, and relative and absolute frequencies, with 95% confidence intervals, depending on the variable described. The diagnostic accuracy of the CAR-OLT score in predicting cardiac contraindications for LT listing was assessed using the area under the receiver operating characteristic (AUROC) curve. The cut-off in the ROC curve was chosen based on the best sensitivity score, as the goal was to exclude true negatives from subsequent diagnostic work-up. Post-LT survival was estimated using Kaplan-Meier analysis. Univariate analysis was performed using univariate logistic regression after analysis of variance (ANOVA) for continuous variables and χ² analysis for categorical variables. Multivariate analysis was conducted using multivariate logistic regression with stepwise forward selection (see supplementary appendix for detailed analysis).

Sample size calculation

The sample size was calculated as follows. The minimum required sample size of 126 subjects was determined using an f2 statistic for multiple regression, with u = 12 (number of predictors), power = 0.8, a significance level of 5%, and an f2 value of 0.15 (mean effect size according to Cohen, 1987). Considering a dropout rate of c = 20% due to adverse cardiological conditions, the sample size increased to nc = n + cn = 151 subjects. Considering an overall dropout rate of d = 45% due to various adverse conditions, the sample size was further increased to nd = n + dn = 183 subjects. Estimation was done using the pwr.f2.test function from the R8 pwr package.

RESULTS

Study population

During the study period, 342 cirrhotic patients were evaluated for LT. The characteristics of the study patients are presented in Table 1 and Supplementary Table 1. Most patients were male (80.7%), with a mean age of 57.5 years, and nearly all were of Caucasian ethnicity (97.4%). Active workers represented 43.6%, and 4.7% had a college-level education or higher. A history of cigarette smoking was reported in 59.7% of patients. The mean BMI was 26.6, and 22.2% of candidates were classified as obese. The most common indications for LT evaluation included alcoholic liver disease (50.9%), hepatocellular carcinoma (46.2%), hepatitis C virus (28.4%), hepatitis B virus (23.4%), and MASH (22.5%). Mean MELD scores at listing and at the time of LT were 16.3 and 18.8, respectively.

Table 1 CAR-OLT score components in our cohort, n (%).

Pretransplant risk factors	Entire cohort (n = 342)	Missing data
Age group		None
< 45	31 (9)
45-49	29 (8.5)
50-54	49 (14.3)
55-59	73 (21.3)
60-64	83 (24.3)
65+	77 (22.5)
Sex
Men	276 (80.7)
Women	66 (19.3)	None
Race
White	333 (97.4)
Black	5 (1.5)
Others	4 (1.1)
Working status
Working for income	149 (43.6)	None
Not working for income	193 (56.4)	None
Education
≤ High school	326 (95.3)
College+	16 (4.7)
Atrial fibrillation
Yes	16 (4.7)
No	326 (95.3)	None
Respiratory failure on ventilator at transplant
Yes	1 (0.3)
No	341 (99.7)	None
Pulmonary hypertension
Yes	11 (3.2)
No	331 (96.8)
Hepatocellular carcinoma
Yes	158 (46.2)
No	184 (53.8)
Hypertension
Yes	118 (34.5)
No	224 (65.5)	None
Diabetes
Yes	128 (37.4)
No	214 (62.6)	None
Heart failure
Yes	16 (4.7)
No	326 (95.3)	None

At pre-transplant imaging, 17.8% of patients had portal vein thrombosis, and 9.5% were treated with enoxaparin. A total of 37.4% had diabetes mellitus, and 22.8% of these received insulin therapy. Arterial hypertension was present in 34.5% of patients, and 19.9% had chronic obstructive pulmonary disease; however, none required oxygen therapy at the time of evaluation. Chronic kidney disease was noted in 13.5% of patients, with 1.7% requiring renal replacement therapy. Only one patient in the cohort was listed in a state of respiratory failure on mechanical ventilation: A 59-year-old woman intubated for pulmonary edema secondary to severe hepatorenal syndrome, which required hemodialysis. A family history of cardiovascular disease was reported in 38.6% of patients, and 15.5% had a personal history of CAD. The individual CAR-OLT score components for our cohort are shown in Table 1, and the distribution of CAR-OLT scores among patients is depicted in Figure 1A.

Figure 1 CAR-OLT score and diagnostic cardiac work-up before liver transplantation. A: Distribution of CAR-OLT scores among patients in the study cohort; B: The CAR-OLT score diagnostic accuracy in predicting cardiac contraindications to listing in our cohort. Diagnostic accuracy indexes are reported for a CAR-OLT score of 23. CAR-OLT: Cardiovascular risk in orthotopic liver transplantation.

Cardiac workup

The results of the cardiac workup are presented in Table 2 and Supplementary Table 2. In the entire cohort, 54 (15.8%) patients were evaluated by exercise electrocardiogram; of these, 9 (20.4%) valid tests showed doubtful or positive results and were referred for coronary angiography. A total of 206 (61.1%) patients underwent stress myocardial perfusion imaging (68% with exercise, 32% with dipyridamole); 68 (32.7%) of these had doubtful or positive results and were subsequently referred for coronary angiography. Fourteen (4.1%) patients were evaluated by dobutamine stress echocardiogram, and only 1 (7.1%) of these had a doubtful result and was referred for coronary angiography. Thirty-four (9.9%) patients underwent coronary CT angiography, and 4 (11.8%) showed significant stenosis, prompting referral for coronary angiography. In total, eighty (23.4%) patients were evaluated by coronary angiography.

Table 2 Cardiac workup and pharmacoeconomic analysis¹, n (%).

Variable	Entire cohort (n = 342)	Patients with CAR-OLT scores ≤ 23 (n = 84)	Potential saving (€)	Potential saving (€) per patient
Exercise electrocardiogram
Tests performed	54/342 (15.8)	18/84 (21.4)	1004.40	11.95
Doubtful/positive result (on valid tests)	9/44 (20.4)	5/13 (38.5)
Stress myocardial perfusion imaging
Tests performed	206/342 (60.2)	58/84 (69)	17864	212.66
Doubtful/positive result	68/206 (33)	12/58 (20.7)
Stress echocardiogram
Test performed	14/342 (4.1)	2/84 (2.4)	140€	1.66€
Doubtful/positive result	1/14 (7.1)	0
Coronary computed tomography angiography
Tests performed	34/342 (9.9)	9/84 (10.7)	1422€	16.93€
Significant stenosis	4/34 (11.8)	1/9 (11.1)
Coronary angiography
Test performed	80/342 (23.4)	13/84 (15.5)	27846€	331.5€
Significant stenosis	25/80 (31.2)	3/13 (23.1)
1 vessel	10	1
2 vessels	9	0
3 vessels	6	2
Myocardial revascularization
Percutaneous coronary intervention	11	1
Coronary artery bypass	1	1
Right heart catheterization	11	1		-
Cardiac contraindications to listing	21/342 (6.1)	2/84 (2.4)
Coronary artery disease	10	1
Valvular heart disease	4	0
Severe pulmonary hypertension	3	1
Atrial fibrillation associated with left ventricular ejection fraction < 45%	4	0
Overall diagnostic work-up	388 diagnostic exams	100 diagnostic exams	48276.40€	574.70€

¹Pharmacoeconomic analysis based on the diagnostic algorithm which defines low-risk patients with CAR-OLT scores ≤ 23, in our cohort 84 patients (24.6%) out of 342.

The cost estimates for each diagnostic test are as follows: Exercise electrocardiogram 55.80 €, myocardial perfusion imaging 308€, stress echocardiogram 70€, coronary computed tomography angiography 158€, coronary angiography 2142€. For exercise electrocardiogram, the test was considered valid if it reached at least 85% of the age-predicted maximum exercise heart rate. CAR-OLT: Cardiovascular risk in orthotopic liver transplantation.

The invasive procedure revealed significant stenosis in 25 (31.2%) patients, 60% of whom had multivessel involvement. Twelve patients underwent coronary artery revascularization before listing, primarily through percutaneous coronary intervention. Eleven (3.2%) patients required right heart catheterization due to a PASP ≥ 40 mmHg on color-doppler echocardiogram. In 5 of these patients, the diagnosis of pulmonary hypertension was not confirmed, and they proceeded with LT evaluation. Three patients were diagnosed with portopulmonary hypertension and were referred to a respiratory specialist for pharmacological treatment, while another 3 were diagnosed with pulmonary hypertension secondary to fluid overload and started on diuretics. After appropriate treatment, these 6 patients were reevaluated by right heart catheterization. Three were subsequently listed for transplantation, while the other 3 were excluded from the list due to failure to achieve a mean pulmonary arterial pressure < 45 mmHg.

A total of 21 patients (6.1%) were deemed to have cardiac contraindications to LT listing (Table 2). The most common reason was severe, non-revascularizable CAD, as determined by the multidisciplinary team (n = 10, 47.6%), followed by valvular heart disease, which included two cases of severe aortic stenosis and two cases of severe aortic insufficiency (n = 4, 19%). Other causes included atrial fibrillation associated with a left ventricular ejection fraction < 45% (n = 4, 19%), and severe pulmonary hypertension despite maximal medical management (n = 3, 14.3%). One patient excluded due to severe pulmonary hypertension also had severe tricuspid regurgitation. Additionally, one transplanted patient with porto-pulmonary hypertension, initially responsive to medical therapy, developed a recurrence of the condition post-LT, which was associated with moderate-to-severe right ventricular dysfunction. Univariate and multivariate analysis of predictors of cardiac contraindication to LT listing are presented in Supplementary Table 3.

CAR-OLT score diagnostic accuracy

Among the 342 candidates, 158 patients (46.2%) received LT. The 1-year post-transplant survival rate was 85% in our cohort. Of the patients who underwent LT, 17 recipients (10.8%) experienced a MACE within the first year after transplant. These results, along with the CAR-OLT score’s ability to predict cardiac MACE one year after LT, have been reported elsewhere[29]. Additionally, detailed data on the post-LT course of patients with significant CAD have been presented in other publication[30].

The diagnostic accuracy of the CAR-OLT score in predicting cardiac contraindications to LT listing is shown in Figure 1B. The CAR-OLT score predicts cardiac contraindications to LT listing with an AUROC of 0.81. The cut-off with the best sensitivity corresponds to a CAR-OLT score of ≤ 23. This cut-off demonstrates 95% sensitivity, 22% specificity, 99% negative predictive value, and 7% positive predictive value for cardiac contraindications to LT listing in our cohort.

In our cohort, 84 patients (24.6%) had a CAR-OLT score ≤ 23 (low-risk candidates). Among these low-risk candidates, only two patients (2.4%) had cardiac contraindications to LT listing. The first was a 53-year-old patient with a CAR-OLT score of 14 and a history of myocardial infarction. Coronary angiography revealed significant three-vessel CAD, which was deemed non-revascularizable by the multidisciplinary team. The second was a 51-year-old patient with a CAR-OLT score of 23, excluded due to severe pulmonary hypertension secondary to excess volume during anesthesia induction, which was unresponsive to subsequent medical therapy. This patient’s screening echocardiogram reported non-measurable PASP.

Table 2 illustrates the potential savings in diagnostic tests if we had listed all candidates with a CAR-OLT score ≤ 23 (low-risk candidates) without subjecting them to additional non-invasive or invasive cardiac diagnostic tests, aside from the electrocardiogram and color Doppler echocardiogram. These 84 patients could have avoided 87 non-invasive stress tests and 13 coronary angiographies, resulting in an estimated total cost saving of €48276.40 for the Italian National Health System. The estimated savings per patient was €574.70.

DISCUSSION

This study contributes to advancing the cardiac evaluation of LT candidates by testing the application of the CAR-OLT score, which was originally designed to predict post-LT MACE in the post-transplant setting, now in the pre-transplant context. The CAR-OLT score, consisting of 12 variables derived from clinical history, was first introduced in 2017[20] to predict MACE one year after LT, focusing primarily on non-ischemic events. It was initially validated in Illinois (United States) on a cohort with a significantly higher cardiovascular risk profile than seen in Italy (e.g., 60% hypertension, 40% obesity, 35% chronic renal failure, and 30% diabetes). However, the CAR-OLT score showed suboptimal performance in predicting 1-year MACE after LT in various cohorts, including those from Pennsylvania (AUROC = 0.613), Rome (AUROC = 0.702), and Turin (AUROC = 0.620)[21,29]. In our study, the diagnostic accuracy of the CAR-OLT score in predicting cardiac contraindications to LT listing was excellent (AUROC = 0.81). We found that a CAR-OLT score cut-off of ≤ 23 effectively identifies a group of low-risk patients who can proceed directly to LT listing without the need for additional provocative testing.

The population in our study closely reflects the demographic and clinical characteristics of LT candidates in Italy. These candidates are predominantly men in their 5th or 6th decade of life, of caucasian ethnicity, with prevalent etiologies including alcohol abuse, hepatitis C virus infection, and a high prevalence of hepatocellular carcinoma. This population exhibits a moderate cardiovascular risk profile, with notable rates of diabetes mellitus (37%), arterial hypertension (34%), and obesity (22%). In this context, proper pre-LT cardiac risk stratification is crucial for ensuring the optimal allocation of organs, which are a limited and invaluable resource.

Optimal cardiac risk estimation in LT candidates remains a topic of debate[31]. Non-invasive stress testing often yields inconclusive results in this population due to specific limitations. Ascites and beta-blocker therapy restrict patients’ ability to exercise and achieve the target heart rate. Inotropic agents, such as dobutamine, have limited efficacy in end-stage liver disease (ESLD) patients because of their impaired chronotropic response. Similarly, vasodilator drugs, including dipyridamole, adenosine, and regadenoson, offer limited utility due to the chronic vasodilation state associated with ESLD[32,33]. In patients undergoing LT evaluation, systematic reviews have reported a positive predictive value of 37% for dobutamine stress echocardiography[34] and 20%–25% for stress myocardial perfusion imaging[35,36]. Coronary CT angiography also shows poor predictive value in this context, as tachycardia occurs in approximately 30% of ESLD patients, contributing to a false-positive rate of about 25%. Furthermore, its use is relatively contraindicated in individuals with kidney disease due to its association with nephrotoxicity[37].

As a result, there is no universal pre-LT cardiological evaluation protocol, and each LT center develops its own diagnostic approach[38]. Given the low sensitivity of non-invasive stress tests in this population, there is a noticeable trend in the literature toward minimizing their use. For instance, the Melbourne protocol stratifies candidates into three risk groups based on factors such as sex, age, diabetes, history of heart disease, and other cardiovascular risk factors. Low-risk candidates undergo a baseline echocardiogram alone, intermediate-risk candidates are referred for stress echocardiography, and high-risk candidates are evaluated with both stress echocardiography and coronary angiography (CT or invasive), depending on stress test results[39]. In contrast, Indiana University adopts a different approach, where LT candidates undergo coronary angiography based on risk factors, even when non-invasive stress test results are negative[40]. Similarly, the Madrid center bypasses non-invasive stress testing entirely for high-risk candidates, assigning them directly to coronary angiography (CT or invasive)[41]. Many experts emphasize the need for a single diagnostic test that can simultaneously provide anatomical and functional information about coronary arteries in LT candidates. In this context, promising preliminary data have emerged on the use of coronary artery calcium scans and coronary CT angiography-derived fractional flow reserve[42,43].

In our study, we identified a low-risk group of patients, such as those with a CAR-OLT score ≤ 23, who exhibited a 99% negative predictive value for cardiac contraindications to LT listing. This cut-off encompassed approximately 25% of the candidates in our cohort. If applied to the study population, it could have eliminated the need for 87 non-invasive diagnostic tests and 13 coronary angiographies, potentially saving an estimated €48000.

Beyond cost savings, an algorithm like ours, which identifies low-risk patients who can safely avoid further cardiologic testing, offers additional benefits. It helps prevent unnecessary complications, such as contrast-induced acute kidney injury associated with non-invasive or invasive coronary angiography, and reduces the risk of false-positive results that would necessitate further investigations. Moreover, this approach shortens the pre-listing evaluation time, facilitating a more efficient transplantation process.

Unfortunately, even for very high scores, it is not possible to establish a CAR-OLT score cut-off above which an LT candidate can be ruled out without further testing. In other words, while the CAR-OLT score is effective in identifying patients at low risk, it cannot reliably identify those at an unacceptably high risk of cardiac contraindications to LT.

The limitations of our study include its retrospective design and the lack of external validation. Validation in other cohorts with different pre-LT cardiac work-ups is necessary before this strategy can be applied on a larger scale. Additionally, as a clinical practice study, our work does not allow for the calculation of sensitivity and specificity values for stress myocardial perfusion imaging in this population, since patients with negative test results did not undergo coronary angiography.

Furthermore, patients with acute liver failure were excluded from our study because they follow a distinct pre-LT cardiac work-up. However, a score such as CAR-OLT could potentially have even greater utility in this population, given the urgency to complete the evaluation process as quickly as possible.

CONCLUSION

In conclusion, within our Italian cohort of patients, the CAR-OLT score proved to be a reliable predictor of cardiac contraindications to LT listing. We propose that a CAR-OLT score cut-off of ≤ 23 identifies a low-risk group of patients who could proceed directly to LT listing without the need for additional provocative cardiac tests. This approach has the potential to reduce barriers to LT listing, lower costs for the National Health Service, shorten waiting times, and alleviate stress for patients.